Anthropic, the AI lab behind the Claude models, has publicly filed its S-1 registration statement with the U.S. Securities and Exchange Commission. The filing is the formal step toward an initial public offering that could rank among the largest in history. Anthropic intends to list its Class A common stock on the Nasdaq Global Select Market under the ticker ANT. The share count and price range are still blank.
The filing gives the public its first audited look at one of the fastest-growing companies ever. Revenue is exploding, and so are the costs of the computing infrastructure behind it. Here is everything we know so far.
Anthropic IPO at a glance
- Ticker and exchange: ANT, Nasdaq Global Select Market
- Lead underwriters: Goldman Sachs and Morgan Stanley, with 22 banks listed on the prospectus cover
- Price range and share count: not yet determined
- Reported valuation: some current and prospective investors value Anthropic at about $1.5 trillion, down from roughly $2 trillion a week earlier.
- Reported timing: the roadshow is expected to begin the week of November 9, per Bloomberg-sourced reports, and Fox Business reported a possible IPO date of November 19 or 20. Anthropic has not confirmed a date.
The key numbers from the S-1
The filing covers the first half of 2026 (January through June), plus full-year figures for 2023 through 2025.
- Revenue reached $16.3 billion in the first half of 2026, up from $1.16 billion in the same period a year earlier, roughly a 14-fold increase. Full-year revenue was $41 million in 2023, $386 million in 2024 and $4.6 billion in 2025. In the second quarter alone, revenue grew 1,368% year over year.
- Net loss was $80.7 billion, compared with about $13 billion in the first half of 2025.
- Operating loss was far smaller: $693 million, down from $2.1 billion a year earlier. On a non-GAAP basis, Anthropic reports adjusted operating income of $1.6 billion, a 10% margin. That figure excludes items such as stock-based compensation, which was $2.2 billion in the half.
- Cash stood at $131.7 billion in cash, equivalents and short-term investments as of June 30. Anthropic says it has raised $141.1 billion in total.
- Customers include more than 11,900 organizations spending at least $100,000 a year, more than 1,900 spending $1 million or more, and about 250 spending $10 million or more. Net revenue retention is about 900%, which means existing customers are spending far more than they did a year ago. Revenue per employee is over $6.6 million.
Why the $80.7 billion loss looks so big
The headline loss is the number most pages are leading with, but it needs context. About $72.4 billion of it is not cash leaving the company. It is a fair-value accounting charge on convertible notes, derivative liabilities and SAFEs (simple agreements for future equity).
The mechanics are straightforward. Those instruments convert into stock, so when Anthropic’s valuation rises, the accounting value of what it owes investors rises with it, and the increase is booked as a loss. The filing says these changes were driven by the rise in Anthropic’s enterprise valuation, including its Series H financing. In plain terms, a higher valuation produces a bigger paper loss.
Strip that charge out and the operating picture looks different. Compute and infrastructure spending fell sharply as a share of revenue, and the operating loss narrowed even as revenue multiplied. The company attributes this to compute efficiency: it says revenue grew 1,368% in the second quarter while compute and infrastructure expenses grew 300%.
One cost line is rising fast, though. Sales, marketing and partnerships expense rose from $120 million in the first half of 2025 to $3.2 billion in the first half of 2026. The filing explains that Anthropic pays fees to the cloud providers that distribute Claude, including AWS, Google Cloud and Microsoft Azure.
The compute bill
Training and running frontier AI models requires enormous amounts of computing power, and the S-1 shows how much Anthropic has committed to buy. The filing states that nearly all of its purchase commitments and leases relate to compute and infrastructure. Some run into the 2040s.
Anthropic will pay SpaceX around $3.1 billion per month for compute powered by NVIDIA GPUs once all promised capacity is delivered. The filing also describes agreements with Microsoft, Nscale, Lambda and other providers. Anthropic says it spreads workloads across AWS Trainium, Google TPUs, NVIDIA GPUs and AMD GPUs.
The total commitment figure is left blank in the version of the filing released today. Earlier reports cited more than $500 billion, and some analysts have tallied higher. Expect the final number to be a focal point when the amended filing arrives.
Governance: a structure with no public precedent
Anthropic describes its capital and governance structure as having “no direct public company precedent,” and the details bear that out.
- Public benefit corporation. Directors must balance shareholder interests, the interests of those affected by the company’s conduct, and a stated public benefit purpose: the responsible development of advanced AI for humanity’s long-term benefit.
- Long-Term Benefit Trust (LTBT). A five-member, financially disinterested trust holds a special Class T share and can elect a majority of the board.
- Founder control. Just before the IPO, Anthropic will issue one Class F share to an LLC owned by its seven co-founders. That share carries 50.1% of the total vote, including the election of directors not chosen by the Trust.
- Limited voice for new investors. Public Class A shareholders get one vote per share but cannot control board elections. Anthropic will be a Nasdaq “controlled company,” which allows it to opt out of some governance requirements, though it says it expects to rely on that only for its nominating and governance committee.
- Public benefit commitments. Anthropic contributed 5% of its fully diluted shares to the Anthropic Public Benefit Fund. The board may issue up to 2% more each time the share price climbs another 50% above prior milestones, starting from the IPO price. Each co-founder has pledged 80% of their personal equity to philanthropy.
For investors, this is a trade-off. The structure is designed to protect Anthropic’s safety mission from short-term market pressure, but it also limits what shareholders can influence.
Risks Anthropic discloses
The risk-factor section runs from page 26 to page 120. Several themes stand out:
- Staying at the frontier. The company says its success depends on continuing to build the most capable models, and that failed or underperforming training runs have happened before.
- Compute dependence. It relies on a small number of cloud and chip partners, some of which are also competitors, investors or customers. The filing acknowledges this can create conflicts of interest.
- Customer concentration. A substantial share of revenue and receivables comes from a limited number of customers and distribution channels.
- Model behavior. Anthropic warns its models may behave in unintended or misaligned ways, including taking harmful actions no user directed. It also notes that using Claude to build the next version of Claude, which it calls “closing the loop,” could let capabilities advance faster than safety measures.
- Self-imposed limits. The filing says it may delay or restrict model releases, as it did with Mythos through Project Glasswing, and that such steps could slow it relative to rivals.
- Growth expectations. Anthropic cautions that past growth rates are not a reliable guide to the future.
Market context: a crowded, nervous IPO window
Anthropic is filing into a shaky market. Reports say that global equities have sold off as investors question whether debt-fueled AI spending will pay off quickly enough, and that higher borrowing costs have added to the caution. Several companies have delayed fall listings.
Two benchmarks loom over the deal. SpaceX priced a $75 billion offering in June that valued it at $1.75 trillion, setting the current IPO record. And reports note the Anthropic offering is expected to become a reference point for how Wall Street values leading AI labs, including rival OpenAI, which is expected to list by early 2027.
What to watch next
- The price range and share count in an amended filing
- Third-quarter revenue, which is blank in the current version of the prospectus
- The total compute commitment, which is also blank
- Roadshow reception and whether valuation talk stays near $1.5 trillion
- Broader AI-stock sentiment, which could move the timetable
What the Anthropic IPO means for businesses building with AI
Strip away the valuation debate and the filing makes one thing clear: enterprise AI adoption is real and accelerating. Thousands of organizations now spend six figures a year on frontier models, and Anthropic’s own prospectus describes AI agents working alongside employees on multi-step tasks. The practical question for most companies is no longer whether to use AI. It is how to integrate it into their products and operations without wasting budget.
That is where a development partner matters. Frontier models are a starting point. Turning them into something that works inside your existing systems, with your data, at a cost you can sustain, takes engineering.
This is the space where Cubix works. Cubix is a software development company with 18+ years of experience, and its services span generative AI, AI chatbots, virtual assistants, NLP models, intelligent automation and recommendation engines, alongside mobile apps, custom software and cloud and MLOps work. For teams watching the Anthropic IPO and wondering how to put AI to work in their own business, Cubix can help.
- Custom AI app development built around a specific business problem
- AI integration into existing software, using APIs from leading model providers
- Generative AI and AI assistants for customer support, internal knowledge and workflow automation
- Model fine-tuning and ongoing optimization as your data and needs change
If you are planning an AI initiative, talk to the Cubix team about where to start.
Frequently asked questions about anthropic AI
1. Can I buy Anthropic stock before the IPO?
Not through normal brokerage accounts. Anthropic is privately held, and its shares are owned by investors and employees. Public trading begins only after the offering is priced and the stock lists.
2. What is the Anthropic ticker symbol?
ANT, on the Nasdaq Global Select Market.
3. Did Anthropic really lose $80 billion?
On paper, yes. The reported net loss for the first half of 2026 is $80.7 billion, but about $72.4 billion is a non-cash accounting charge tied to its rising valuation. The operating loss was $693 million.
4. When will the Anthropic IPO happen?
No date is confirmed. Reports point to a roadshow starting around November 9 and a pricing window in late November, but those timelines can shift with market conditions.
5. What is the IPO price?
Not yet set. The price range will appear in an amended S-1.
